Dietary exposures are believed to contribute to the majority of human cancers although most of the specific causal factors have not yet been identified. Since diet is a universal and modifiable exposure, research in this area offers the opportunity of significantly reducing cancer incidence and mortality. The Nutritional Epidemiology Branch conducts independent and collaborative research on the role of diet and nutritional status in cancer etiology. The Branch seeks to generate and test specific nutritional hypotheses, providing a scientific basis for public health recommendations and further understanding of the molecular basis of carcinogenesis. The Branch currently focuses on four broad exposure-drivenhypotheses involving: 1) energy balance (obesity and physical activity); 2) hyperinsulinemia and insulin resistance (taken broadly to include glycemic index and the IGF axis); 3) one-carbon metabolism; 4) meat, including potentially carcinogenic products of processing and cooking. The Branch also investigates other macronutrients and foods (fat, fiber, fruits and vegetables,) as well as micronutrients (vitamins D, E; calcium; selenium; carotenoids). Branch investigations in these priority areas use data on diet, supplement use, body size, physical activity, and blood- and tissue-based analytes. Many of these studies are also incorporate DNA-based genetic information. Because of the extensive biological research underlying contemporary nutrition, we frequently develop multidisciplinary studies with metabolic and/or molecular components. The role of nutrition is currently being evaluated in studies of lung, esophageal, stomach, colorectal, breast, endometrial, cervical, and prostate cancers as well as non-Hodgkin's lymphoma. Research approaches include descriptive analyses to generate hypotheses, analytic cohort and case-control investigations (with increasing emphasis on large prospective studies), nutritional intervention studies (clinical trials), metabolic studies, and biologic marker and genetic susceptibility projects. The Branch research program is designed to overcome several scientific obstacles to advancing the field: 1) Exposure assessment error. NEB, in conjunction with other NCI Divisions and other investigators around the world, has been conducting a series of methodologic studies to determine if we are missing key nutrition-cancer relations because of food frequency questionnaire measurement error and whether we need to use more intensive (and expensive) but more accurate assessment instruments. NEB is also expanding its work in nutrition-gene interactions. As we work through the problem of gene/SNP/haplotype selection in the face of complex biochemical pathways, NEB?s expanding work in nutrition-gene interactions can reveal ?sharpened? relative risks among the genetically susceptible (e.g., those with a particular allelic variant of a metabolizing enzyme-encoding gene) and thereby clarify the role of particular nutritional factors (alone or as part of complex dietary mixtures). 2) Inadequate exposure range. NEB?s investigations include existing (AARP, PLCO) and proposed (India) cohorts with wide reported intakes ranges for several key dietary factors. 3) Confounding. Relative risks for many nutritional factors are likely to be modest and easily confounded, even with adjustment for covariates. NEB approaches to this problem include: a) conducting or collaborating on randomized controlled trials, e.g., Polyp Prevention Trial, SELECT; b) using ?Mendelian randomization?, a strategy based on the premise that an association between an allelic variant mimicking, e.g., low micronutrient exposure is less likely to be biased by confounding (and measurement error) than an association with the dietary factor itself.